Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sdd.hp.com!elroy.jpl.nasa.gov!decwrl!deccrl!bloom-beacon!eru!hagbard!sunic!news.funet.fi!polaris.utu.fi!magi From: magi@utu.fi (Marko Gronroos) Newsgroups: comp.ai.philosophy Subject: Re: Continuous vs discrete Message-ID: Date: 2 Apr 91 18:38:09 GMT References: <91082.223501DOCTORJ@SLACVM.SLAC.STANFORD.EDU> <1991Mar25.141743.21124@news.larc.nasa.gov> <19175@ogicse.ogi.edu> <19392@ogicse.ogi.edu> Sender: usenet@polaris.utu.fi (Usenet News) Organization: University of Turku, Finland Lines: 85 In-Reply-To: maxwebb@ogicse.ogi.edu's message of 30 Mar 91 20:55:38 GMT > I posted an earlier version of this article in which i was > more heated and sarcastic than I needed to be. Sorry. Canceled it, > but it might have got out anyway. Sorry for giving any reason. Few years ago I made a rule never to write news articles after 10 PM.. It seems that I forgot it again.. > Marko, it all comes down to this. There are simulations of > neural nets which are close enough to the biological version to > reproduce _waveforms_. Example: predator evasion reflex of > tritonia. Example: swimming behavior of the Lamprey. Example: > stomach ganglia of the lobster. You want ref's? Yes, if it isn't too much trouble. About those simulations, not examples. I suppose the simulations were done in at least virtually continuous time? I am not saying that there are no such simulations, I'm just saying that they are rare, and _most_ current ANN models don't care much about the biological version. People seem to like speed more than correct logical behaviour. Are we taking a false path from our way of creating a thinking machine if we try to optimize it too much before it even works? > Since these all do work well, doesn't that kind of indicate > that these networks are simulable? Hmm. The differences may not cause too big errors in simple simulations such as these, but may cause big errors in networks of larger size and logical complexity. > >> it is not computing it to 32 bits accuracy! > >32 bits? Integer? Not floating point? Sorry again. Thought that you were talking about int's... But I'd say that my example would have been good if you had been talking about int's. > Yeah, Yeah, of course use fp. But does floating point rep. > stop being discrete? surprise to me. > Weren't we talking about discrete vs. continuous? Yes, floats are discrete, and I agree with you in that 32 bit floats can represent "virtually continuous" quantities. Some people in this newsgroup say that even space and time could be discrete. The question number 1 is that at what point of discretity would the difference between the simulation and the nature become significant? (With ANN's..) > Another cheap shot. You seem to be more interested in scoring > rhetorical points than answering the content of my posting. Another cheap shot. You seem to be more interested in scoring rhetorical points than answering the content of my posting. :-) Sorry.. Ok, no more cheap shots.. > Watch your tongue. These machines manage to simulate lot's of > time/space continuous systems (turbulent air flow, physics > of vlsi device geometries). Why are you so pessimistic in this > particular arena? In those systems that you mentioned simulations try to simulate real physical behaviour. Most current artificial neural networks simulate the presumed logical behaviour of some objects that we really don't know much about. Most of those simulations are made in virtually continuous quantity and space and time. Neural nets of the brain would be too complex to simulate in that way with conventional computers, wouldn't they? > If you spent more time reading the literature, and less time > speculating about problems that _might_ or _might_not_ exist > you might be less pessimistic about our chances of success. Still remember the conventional AI and it's failure to produce thinking machines? I could go and read tons of conv. AI literature while someone who is pessimistic about it could do the trick with ANN's. Why choose the first road that comes in my sight and travel in a false direction, if I can be pessimistic for a little time, and then take the most propably correct road? It's not only important that one studies, it's also important _what_ one studies. Can you give me some hints? :-) ------------------------------------------------------------------------------- Marko Gronroos ! Tel. +358-21-445613 ! Karvataskunkatu 10 H 100 ! magi@utu.fi ! Computer Scientists do it 20610 Turku ! ! with bigger hardware. Finland ! ! ------------------------------------------------------------------------------